Researchers in the Netherlands have discovered that RFID systems, intended for tracking hospital kit, can fatally interfere with life-support systems from a distance of 30 centimetres.

The tests were pretty rigorous: 41 different devices from 22 manufacturers were each tested against passive and active RFID systems. The passive system, operating at 868MHz, generated problems in 26 devices, while the active system, sitting at 125kHz, only affected 8 of them.

It's not entirely surprising that passive tags generate more interference, as such systems require the reader to generate an electromagnetic field that can be used to power the tag. But it is surprising that the tags generate any interference at all, since they are supposed to be entirely benign to other electronics, and one would hope that critical equipment would be adequately shielded.

The research is to appear in the Journal of the American Medical Association in an editorial, written by the president of the Institute for Healthcare Improvement Dr. Donald Berwick, which calls for immediate action:

"Attention must be paid to these disturbing findings ... It seems that hospitals, regulators, and manufacturers certainly have some immediate work to do including examining whether similar problems are occurring in hospital critical care units".

There have been no recorded incidents of real-world interference, but no-one has been looking for them and the editorial goes on to recommend hospitals immediately start monitoring for suspected incidents.

No patients were connected to the equipment during the research, so there's been no death-by-RFID yet, but the use of tags to track equipment around the hospital is growing rapidly and without much consideration for the impact on other electronics.

Hospitals will need to evaluate the research, and probably do more of their own. The solution may be as simple as not putting RFID readers in intensive care departments, or asking manufacturers to increase shielding, but it certainly demands more study. ®

25 comments posted — Comment period finished